Thin film transistor redundancy structure

Abstract

In a microcircuit device such as a memory chip, where a bank of state devices such as fuses and anti-fuses determine the enabling and disabling of redundant circuitry, a scheme for blowing one or more state devices by applying a programing voltage through a switching circuit comprising thin film transistors (TFTs) which are not damaged by the device blowing, programming voltage. The TFTs can be activated by a low voltage enable signal provided by a state device designator logic module.

Description

This is a continuation of application Ser. No. 08/313,627, filed Sep. 27, 1994, now U.S. Pat. No. 5,552,743.

FIELD OF THE INVENTION

This invention relates to semiconductor integrated circuits and, more specifically, to the use of state devices such as fuses and anti-fuses to disable a defective portion or block of an integrated circuit and enable a redundant block.

BACKGROUND OF THE INVENTION

Semiconductor integrated circuits contain large numbers of electronic components such as capacitors, diodes, resistors and transistors built on a single chip. Due to the microscopic scale of these circuits, they are susceptible to component defects caused by material impurities and fabrication hazards.

In order to circumvent this problem, chips are built with redundant components and/or circuits that can be switched-in in lieu of corresponding circuits found defective during testing. Usually the switching-out of a defective component or circuit and the switching-in of a corresponding redundant component or circuit is accomplished by using program logic circuits which are activated by altering or programming certain state devices such as fuses or anti-fuses built into the chip circuitry. The process of altering or programing a subset of the available fuses or anti-fuses in order to switch-in the redundant circuitry is also known collectively as programming or repairing.

FIG. 1 illustrates a typical prior art structure for programming a random access memory (RAM) type integrated circuit (IC). The structure involves a bank 1 of state device modules which control the programming logic circuitry 2, which actually makes the switch between defective and redundant circuit portions in the RAM array 3 based on the output of the state device memory bank 1. The bank contains a number of state device circuit modules. Each module contains at least one state device.

State devices are conductivity alterable component. They are manufactured in an initial state, either open (very low conductivity) or closed (high conductivity), and can be altered or programmed to assume the opposite state. Normally, this alteration is only one way, i.e. an altered device cannot be returned to its previous state. Fuses are manufactured closed and are blown open by applying a sufficient current through the device such that resistive heating causes the normally conductive fuse element to melt or explode, thereby forming an open circuit. Anti-fuses such as dielectric capacitors are manufactured open and are blown or programmed to the closed state by applying a sufficient voltage above their breakdown voltage across their terminals. At this voltage the dielectric layer separating the conductive plates of the capacitor ceases to be electrically insulating. The capacitor then permanently forms a closed circuit between its two terminals.

Although many different state device module circuit designs are possible, each must perform two basic functions: First, altering the state device without subjecting the programming logic circuitry to the programming voltage or current, and second, communicating the state of the fuse or anti-fuse to the programming logic circuitry. FIG. 2 shows a simplified state device module circuit containing one fuse-type state device, a fuse 4 which is connected to ground 5 and a node 6. The node is connected through a resistor 7 to output terminal V_out 8. The node is also connected to terminal V_in 9 and a conductive pad 10.

During programming, a high voltage V_bd is applied to the pad using a probe. The current produced by this voltage causes the fuse to blow, creating an open circuit. V_out is protected from V_bd by resistor 7. After programming is complete, V_bd is removed from the pad.

During operation, V_cc is applied to the V_in terminal. If the fuse is open, V_out will show a positive voltage as seen through the resistor. If the fuse is closed, V_out will show ground.

Other circuitry may be provided to limit the current through the fuse during operation. In the circuit of FIG. 2, when the fuse is not blown, a direct path exists from V_in to ground. In the simplest case, a resistor or some other current limiting device placed between the node 6 and V_in would provide this function.

Typically, the voltage required to alter a fuse or other state device is high enough to damage the transistors and capacitors which make up much of the programming logic circuitry and memory arrays on a typical integrated circuit microchip. The high voltage cannot be multiplexed through other circuitry since it would damage the circuitry typically used for this purpose. In order to protect this circuitry, each state device is given a dedicated pad as a point where the high voltage can be exclusively applied, thus avoiding application of this voltage to the sensitive circuits. Since the pads must be large enough to be engaged by a probe, they take up a substantial portion of the available space on each microchip. This limits the number of pads and state devices. In turn, chip designers have devised elaborate schemes and structures to offer the maximum redundancy with the minimum number of state devices.

In addition, it is difficult to develop redundancy structures whereby a single state device in a bank of state devices may be selectively altered using logic circuitry. The problem here is that the transistors and other devices used to make up the logic circuitry are themselves susceptible to failure when directing voltages and currents high enough to alter the state device.

Because of the use of dedicated altering pads, the current prior art solution requires programming prior to packaging. Since the programming pads are covered with passivation prior to assembly, any switch to redundant circuitry must occur before burn-in and thus, delivery to the customer. Currently, failures detected during and after burn-in result in the entire chip being scrapped.

It would be desirable, therefore, to have a state device module which is capable of receiving a low voltage programming signal which allows high voltage to alter the state device without that high voltage being carried through to other circuitry on the chip, both during and after the programming voltage is applied. It is also desirable to be able to perform this programming before and/or after packaging the integrated circuit.

SUMMARY OF THE INVENTION

The principal and secondary objects of this invention are to provide a structure whereby sensitive microcircuit components may be protected from incompatible voltages required by an element in the microcircuitry. It is a further object of this invention to provide a structure that requires fewer pads on an integrated circuit chip solely dedicated to altering or programming a finite number of state devices. It is another object of this invention to create a structure whereby the state devices may be altered before and/or after burn-in, packaging and delivery to the customer.

These and other objects are achieved by adding a state device designator logic module which selectively sends an enable signal to those state device modules having a state device to be altered. Each state device module, in turn, uses switching devices such as thin film transistors (TFTs) capable of withstanding high voltages to apply the high programming voltage or current to the state device element while protecting downstream circuitry to which the module is connected.

A TFT is a form of a field effect transistor having a control gate and two current carrying terminals, namely a source and a drain. TFT gate dielectrics are typically thicker than bulk MOS gate dielectrics, and can therefore withstand the high programming voltage.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a prior art block diagram of a typical memory redundancy scheme;

FIG. 2 illustrates a prior art bank of state device module circuits;

FIG. 3 illustrates a bank of state device modules connected to a state device designator logic module according to the invention;

FIG. 4 illustrates a state device module circuit according to the invention;

FIG. 5 illustrates an alternate embodiment of a state device module circuit according to the invention; and

FIG. 6 shows a plan view of a semiconductor layout structure for implementing the circuit of FIG. 3.

DESCRIPTION OF THE PREFERRED EMBODIMENT OF THE INVENTION

Referring now to the drawings, there is shown in FIG. 3 a state device designator logic module 15 which uses as its input address lines 16 carrying voltage levels corresponding to memory locations on a memory array. When programming is enabled by a particular value on the programming enable line 17, the particular address arriving on the address lines is interpreted by bulk CMOS logic within the state device designator module to select one of N input lines 18 for programming. Each input line is connected to one of N state device modules 19,20,21 and 22. Each state device module contains a circuit which comprises at least one element which is a state device. In this embodiment, the element is a bulk CMOS capacitor (23 in FIG. 4). The additional circuitry in each state device module serves to allow for altering or programming the capacitor without subjecting the output terminal V_out 24 to the high programming voltage. This additional circuitry is comprised of switching elements which are not susceptible to damage at the applied programming voltage. One type of circuit element that satisfies this condition is thin film transistors, which are typically used as load elements in SRAM memory cells.

Since V_out is wired to other sensitive microcircuit components, V_out must be isolated from voltages such as the programming voltage, which are incompatible with those components. Incompatible voltages would be those greater than the dielectric breakdown voltages of the components. The circuitry in each state device module must also protect the capacitor from the programming voltage if it is not to be programmed. Finally, the circuitry must be able to sense the state of the capacitor during normal operation of the IC and to supply that information via a signal on V_out.

FIG. 4 shows the circuitry of a single state device module 19. During programming, a first supply voltage input terminal V_cc 25 is held at a high voltage which is typically +5 volts, while a second supply voltage input terminal V_low 26 is held at -15 to -20 volts. V_low is therefore opposite in polarity to that of V_cc. Each of these voltages can be supplied by a single pad or pin servicing the entire IC. The program enable input terminal V_in 27 is biased high at +5 volts, which turns off thin film transistor T₂. This in turn brings the gate 28 of T₁ low (-15 to -20 volts) through resistor 29, the other side of which is tied to V_low 26. With its gate low, T₁ is turned on, pulling node A 23a to the V_low voltage. A bulk CMOS capacitor 23 with a dielectric thickness of around 110 angstroms can only withstand about 12 to 15 volts before breaking down. The bulk CMOS capacitor is used in order to simplify processing, since the capacitor can be formed at the same time as other bulk devices. The capacitor dielectric would then be the same thickness as gate dielectric layers used by the bulk devices. With the differential voltage (V_cc -V_low) of 20 to 25 volts placed across its terminals, the capacitor quickly blows, forming a closed circuit between V_cc and node A. This differential voltage fulfills the role of a programming voltage. When programming is complete, V_low is released and V_in remains forced high to +5 volts by the state device designator to keep T₂ turned off. T₁ remains on but the conductance of the blown capacitor is higher than or comparable to that of T₁. Therefore, node A is pulled towards V_cc.

Since the voltage at node A is between V_cc and V_ss, T₄ has a non-zero voltage applied across its source and drain terminals. The gate of T₄ is tied to V_ss, turning T₄ on, which pulls V_out 24 toward V_cc along with node A. V_out is used by the programming logic (2 in FIG. 1) to effect repairing of the IC.

To prevent the capacitor from being blown during programming of other modules, when V_low is held at -15 to -20 volts, V_in is merely biased low at 0 volts. In this case, T₂ is on, allowing the gate 28 of T₁ to be pulled high through T₂ to +5 volts appearing on V_cc. This keeps T₁ shut off, isolating the capacitor 23 from V_low. Again, when programming is complete, V_low is released and V_in is forced high to +5 volts. This turns off T₂. The gate 28 of T₁ is pulled low to the voltage appearing on the terminal V_ss 26a through the resistor 29 and T₃, thereby turning T₁ on. Since the capacitor is intact and does not conduct current at 5 volts, node A is pulled toward V_ss through T₁ and T₃. In this circuit, T₃ has its gate wired to its drain to form the cathode of a TFT diode which is backwired (cathode to minus and anode to plus) between V_ss and V_low.

T₄ acts as an isolation element to prevent the voltage on V_low from reaching V_out 24 during programming, thereby protecting the bulk MOS devices in the programming logic (2 in FIG. 1), T₄ 's gate is wired to V_ss. When the state device module is selected by the designator to blow the capacitor, node A 23a is pulled to V_low as previously described. Once node A is equal to or less than the gate potential of T₄ (V_ss), T₄ turns off, thereby protecting the bulk device, wired to V_out from node A and hence, V_low. Once V_low is released and normal, "sensing" operation of the circuit resumes, and node A is either pulled towards V_cc or V_ss depending on whether the capacitor is intact or not. In either case, since the gate of T₄ is biased at V_ss, T₄ is on and passes the voltage on node A to V_out and subsequent circuitry. Therefore, the voltage appearing on V_ss is, in essence, an element status interrogating signal.

In this embodiment, TFTs are used as the circuit elements in the state device module. TFTs typically provide a low amount of current when turned off. This current is insufficient to cause breakdown of bulk gate dielectrics, but is sufficient to pass a voltage if no higher current source is provided. Therefore, when node A and T₄ 's gate are both at V_ss, V_out will eventually be pulled completely to V_ss as well. The circuitry to which V_out is connected (in the programming logic circuitry) is typically a bulk device gate terminal, and therefore does not provide a DC source of current that can overcome the "leakage" off current of T₄.

An alternate embodiment of the state device module circuitry is shown in FIG. 5. Here, the isolating circuitry has shifted to the V_cc side of the module 49. As with the previous embodiment, during programming, V_cc 50 is held at a high voltage, +5 volts, while V_low 51 is held at -15 to -20 volts. However, program enable input line V_in 52 is biased low to turn on thin film transistor T₂. This brings the gate 53 of T_l low through T₂. With its gate low, T₁ is turned on, pulling node A 54 to the V_cc voltage. This places the differential programming voltage of 20 to 25 volts (V_cc -V_low) across the capacitor 55, which causes it to blow, forming a closed circuit between V_low and node A.

To prevent the capacitor from being blown during programming when V_low is held at -15 to -20 volts, V_in is biased high. This causes T₂ to be off, allowing the gate 53 of T₁ to be pulled high through the resistor 56, thereby turning T₁ off as well. This isolates the capacitor 55 from V_cc and keeps it from being blown.

During sensing, V_low is released and V_in held low, turning on T₂. The gate 53 of T₁ is pulled to V_ss through T₃ and T₂, thereby turning T₁ on. If the capacitor is intact and does not conduct current at 5 volts, node A is pulled to V_cc through T₁. If the capacitor is blown into a closed circuit, node A is pulled to V_ss through T₃. Again, T₄ acts as an isolation element to prevent the programming voltage from reaching V_out 57 and bulk MOS devices connected to V_out.

The invention allows for a reduction in the number of required pads since each state device module no longer requires a dedicated pad for programming. The separate input to each state device module (V_in) is in the form of a programming enable line, not a programming pad. The low programming voltage may be applied at a single pad which is wired to V_low of all the state device modules. Pads will still be required for V_cc and other IC functions such as the addressing pads. However, these pads are required in any case as part of the standard chip configuration. The only additional pad required by the invention is one for V_low.

Programming may be done in the traditional manner through probe while the individual IC's are still united in a wafer. Or, since programming is done through pads which will be connected to pins when the IC is packaged, programing may be performed after packaging, burn-in or delivery to the customer. In this way, failures after burn-in may now be recovered, reducing scrap.

It shall be understood that the state device designator logic module need not have individual address lines running into it. Addressing data can arrive through other means such as one or more multiplexed signal lines. A programming enable pad may not be necessary if the module is designed to accept an enable message through the address line or lines.

Other state devices may be used such as resistors, polysilicon fuses or other anti-fuse devices such as interpoly capacitors, intermetal capacitors and those using. sacrificial interfacial oxides without departing from the invention. Circuitry using these devices can make similar use of TFTs to protect bulk MOS logic circuits from the high programming voltages.

If the invention is implemented on an SRAM chip already Using thin film transistors as memory cell load elements, no additional fabrication processing is required since both the load elements and the state device modules can be created during the same fabrication steps. In any integrated circuit which fabricates separate TFTs in other circuitry for purposes other than the redundancy scheme and not part of any means for isolating the programing voltages from more sensitive components on the microcircuit chip, the invention TFTs can be fabricated during those steps which fabricate the separate TFTs.

An embodiment of an IC structural layout for the fuse module circuitry is shown in FIG. 6. The circuit is formed on an n-well which is wired to V_cc 30. The capacitor 31 is formed between the n-well and a conductive polysilicon structure 32 separated from the n-well by a bulk gate oxide layer. A first terminal of the capacitor is wired through the n-well to a first p-type source/drain electrode 33 of a thin film field effect transistor, T₂. T₂ 's gate electrode 34 is a conductive polysilicon structure which is connected to V_in through a contact point 35. T₂ 's second source/drain electrode 36 is wired via a conductive polysilicon structure 37 to the gate electrode 38 of another thin film field effect transistor, T₁. T₁ 's first source/drain electrode 39 is connected via another conductive polysilicon structure 40, forming node A, to the capacitor's second terminal 32. T₁ 's first source/drain electrode 39 also forms the first source/drain electrode of another TFT, T₄. T₄ 's gate 42 is connected to V_ss through a contact 43, and T₄ 's second source/drain electrode is connected through a contact 44 to V_out. A high value resistor 45 connects T₁ 's second source/drain 46 to T₂ 's second source/drain 36. T₁ 's second source/drain is also wired to V_low through a contact 47.

The layout of FIG. 6 is only one example. Many other structural configurations are available without departing from the invention. In addition, the arrangement and makeup of the circuit elements may be altered without departing from the invention.

While the preferred embodiments of the invention have been described, modifications can be made and other embodiments may be devised without departing from the spirit of the invention and the scope of the appended claims.

Claims

1. A semiconductor device, comprising:

designator circuitry that selectively provides a designation signal having a particular condition;

microcircuit components;

a state device that changes from a first state to a second state when subjected to a programming voltage differential; and

control circuitry that subjects the state device to the programming voltage differential when the designator circuitry provides the designation signal having the particular condition to the control circuitry, the control circuitry isolating the microcircuit components from voltages associated with the programming voltage differential that are incompatible with the microcircuit components, and the control circuitry providing a signal to the microcircuit components that is indicative of the state of the state device.

2. The device of claim 1, wherein the control circuitry includes thin film transistors that withstand voltages associated with the programming voltage differential that are incompatible with the microcircuit components.

3. The device of claim 1, wherein the state device is formed with the microcircuit components.

4. The device of claim 1, wherein the state device is a fuse.

5. The device of claim 1, wherein the state device is an antifuse.

6. The device of claim 1, whereto the control circuitry includes an isolating transistor a gate of which receives an element status interrogating signal.

7. The device of claim 1, wherein the designator circuitry selectively provides additional designation signals at least one of which has the particular condition, and further comprising additional microcircuit components, state devices, and control circuitry, and wherein the additional control circuitry subjects the additional state devices to the programming voltage differential upon receipt of the additional designation signals having the particular conditions, and the additional control circuitry isolates respective ones of the additional microcircuit components from voltages associated with the programming voltage that are incompatible with the additional microcircuit components, and the additional control circuitry provides signals to the additional microcircuit components that are indicative of the respective states of the additional state devices.

8. The device of claim 7, wherein dependent the conditions of the designation signals, some of the state devices may be changed from a first to a second state while others of the state devices are not changed.

9. The device of claim 7, wherein creation of the programming voltage differentials for each control circuit is caused in part through a single low voltage signal.

10. The device of claim 7, wherein the control circuitry and the additional control circuitry each share a common node and creation of the programming voltage differentials is caused in part through a low voltage signal at the common node.

11. The device of claim 1, wherein the particular condition is a first condition and the designator circuitry also selectively provides the designation signal having a second condition, and wherein if the designator circuitry provides the designation signal having the second condition, the control circuitry does not subject the state device to the programming voltage differential.

12. The device of claim 1, wherein if the designation signal does not have the particular condition, the control circuitry does not subject the state device to the programming voltage differential.

13. The device of claim 1, wherein the particular condition is a logic high voltage.

14. The device of claim 1, wherein the control circuitry controls the state of at least one additional state device.

15. The device of claim 1, wherein the microcircuit components are dynamic memory.

16. The device of claim 1, wherein the semiconductor device is on a wafer.

17. The device of claim 1, further comprising packaging and a lead frame.

18. A semiconductor device, comprising:

designator circuitry that selectively provides a plurality of designation signals;

a plurality of microcircuit components;

a plurality of state device modules each including:

a state device that changes from a first state to a second state when subjected to a programing voltage differential; and

control circuitry that subjects the state device to the programming voltage differential if the designation signal having a particular condition is provided by the designator circuitry to the control circuitry, the control circuitry isolating one of the plurality of microcircuit components from voltages associated with the programming voltage that are incompatible with the microcircuit components, and the control circuitry providing a signal that is responsive to the state of the state device to one of the microcircuit components.

19. The device of claim 18, wherein the particular condition is a high voltage.

20. The device of claim 18, wherein the particular condition is the same for each control circuit.

21. The device of claim 18, wherein the control circuitry of each of the plurality of state device modules shares a common node and wherein creation of the programming voltage is caused in part by a low voltage at the common node.

22. The device of claim 18, wherein the control circuitry of each of the plurality of state device modules includes thin film transistors that withstand voltages associated with the programming voltage differential that are incompatible with the microcircuit components.

23. The device of claim 18, wherein the control circuitry of each of the plurality of state device modules includes an isolating transistor a gate of which receives an element status interrogating signal.

24. The device of claim 18, wherein the particular condition is a first condition and the designator circuitry also selectively provides the designation signal having a second condition, and wherein if the designator circuitry provides the designation signal having the second condition, each of the control circuitry do not subject the state device to the programming signal.

25. The device of claim 18, wherein the control circuitry of each of the plurality of state devices that receives one of the designation signals that does not have the particular condition does not subject a state device to the programming voltage differential.

26. The device of claim 18, wherein the semiconductor device is on a wafer.

27. The device of claim 18, further including packaging and a lead frame.

28. A method for controlling and providing an indication of a state of a state device that changes from a first state to a second state when subjected to a programming voltage differential, the method comprising the steps of:

providing a designation signal having a particular condition when it is desired to change the state of the state device;

subjecting the state device to the programming voltage differential when the designation signal having the particular condition is provided;

isolating microcircuit components from voltages associated with the programming voltage that are incompatible with the microcircuit components; and

providing a signal to the microcircuit components that is indicative of the state of the state device.

29. The method of claim 28, further including the steps of:

providing additional designation signals at least one of which has the particular condition;

subjecting additional state devices to the programming voltage differential when an associated designation signal having the particular condition is provided;

isolating additional microcircuit components from voltages associated with the programming voltage that are incompatible with the microcircuit components; and providing signals to the microcircuit components that are indicative of the states of the additional state devices, respectively.